Dynamic Programming Approaches for Estimating and Applying Large-scale Discrete Choice Models

People go through their life making all kinds of decisions, and some of these decisions affect their demand for transportation, for example, their choices of where to live and where to work, how and when to travel and which route to take. Transport related choices are typically time dependent and characterized by large number of alternatives that can be spatially correlated. This thesis deals with models that can be used to analyze and predict discrete choices in large-scale networks. The proposed models and methods are highly relevant for, but not limited to, transport applications. We model decisions as sequences of choices within the dynamic discrete choice framework, also known as parametric Markov decision processes. Such models are known to be difficult to estimate and to apply to make predictions because dynamic programming problems need to be solved in order to compute choice probabilities. In this thesis we show that it is possible to explore the network structure and the flexibility of dynamic programming so that the dynamic discrete choice modeling approach is not only useful to model time dependent choices, but also makes it easier to model large-scale static choices. The thesis consists of seven articles containing a number of models and methods for estimating, applying and testing large-scale discrete choice models. In the following we group the contributions under three themes: route choice modeling, large-scale multivariate extreme value (MEV) model estimation and nonlinear optimization algorithms. Five articles are related to route choice modeling. We propose different dynamic discrete choice models that allow paths to be correlated based on the MEV and mixed logit models. The resulting route choice models become expensive to estimate and we deal with this challenge by proposing innovative methods that allow to reduce the estimation cost. For example, we propose a decomposition method that not only opens up for possibility of mixing, but also speeds up the estimation for simple logit models, which has implications also for traffic simulation. Moreover, we compare the utility maximization and regret minimization decision rules, and we propose a misspecification test for logit-based route choice models. The second theme is related to the estimation of static discrete choice models with large choice sets. We establish that a class of MEV models can be reformulated as dynamic discrete choice models on the networks of correlation structures. These dynamic models can then be estimated quickly using dynamic programming techniques and an efficient nonlinear optimization algorithm. Finally, the third theme focuses on structured quasi-Newton techniques for estimating discrete choice models by maximum likelihood. We examine and adapt switching methods that can be easily integrated into usual optimization algorithms (line search and trust region) to accelerate the estimation process. The proposed dynamic discrete choice models and estimation methods can be used in various discrete choice applications. In the area of big data analytics, models that can deal with large choice sets and sequential choices are important. Our research can therefore be of interest in various demand analysis applications (predictive analytics) or can be integrated with optimization models (prescriptive analytics). Furthermore, our studies indicate the potential of dynamic programming techniques in this context, even for static models, which opens up a variety of future research directions.

Dynamic Programming Approaches for Estimating and Applying Large-scale Discrete Choice Models

People go through their life making all kinds of decisions, and some of these decisions affect their demand for transportation, for example, their choices of where to live and where to work, how and when to travel and which route to take. Transport related choices are typically time dependent and characterized by large number of alternatives that can be spatially correlated. This thesis deals with models that can be used to analyze and predict discrete choices in large-scale networks. The proposed models and methods are highly relevant for, but not limited to, transport applications. We model decisions as sequences of choices within the dynamic discrete choice framework, also known as parametric Markov decision processes. Such models are known to be difficult to estimate and to apply to make predictions because dynamic programming problems need to be solved in order to compute choice probabilities. In this thesis we show that it is possible to explore the network structure and the flexibility of dynamic programming so that the dynamic discrete choice modeling approach is not only useful to model time dependent choices, but also makes it easier to model large-scale static choices. The thesis consists of seven articles containing a number of models and methods for estimating, applying and testing large-scale discrete choice models. In the following we group the contributions under three themes: route choice modeling, large-scale multivariate extreme value (MEV) model estimation and nonlinear optimization algorithms. Five articles are related to route choice modeling. We propose different dynamic discrete choice models that allow paths to be correlated based on the MEV and mixed logit models. The resulting route choice models become expensive to estimate and we deal with this challenge by proposing innovative methods that allow to reduce the estimation cost. For example, we propose a decomposition method that not only opens up for possibility of mixing, but also speeds up the estimation for simple logit models, which has implications also for traffic simulation. Moreover, we compare the utility maximization and regret minimization decision rules, and we propose a misspecification test for logit-based route choice models. The second theme is related to the estimation of static discrete choice models with large choice sets. We establish that a class of MEV models can be reformulated as dynamic discrete choice models on the networks of correlation structures. These dynamic models can then be estimated quickly using dynamic programming techniques and an efficient nonlinear optimization algorithm. Finally, the third theme focuses on structured quasi-Newton techniques for estimating discrete choice models by maximum likelihood. We examine and adapt switching methods that can be easily integrated into usual optimization algorithms (line search and trust region) to accelerate the estimation process. The proposed dynamic discrete choice models and estimation methods can be used in various discrete choice applications. In the area of big data analytics, models that can deal with large choice sets and sequential choices are important. Our research can therefore be of interest in various demand analysis applications (predictive analytics) or can be integrated with optimization models (prescriptive analytics). Furthermore, our studies indicate the potential of dynamic programming techniques in this context, even for static models, which opens up a variety of future research directions.

Constraint Programming-based Job Dispatching for Modern HPC Applications

A High-Performance Computing job dispatcher is a critical software that assigns the finite computing resources to submitted jobs. This resource assignment over time is known as the on-line job dispatching problem in HPC systems. The fact the problem is on-line means that solutions must be computed in real-time, and their required time cannot exceed some threshold to do not affect the normal system functioning. In addition, a job dispatcher must deal with a lot of uncertainty: submission times, the number of requested resources, and duration of jobs. Heuristic-based techniques have been broadly used in HPC systems, at the cost of achieving (sub-)optimal solutions in a short time. However, the scheduling and resource allocation components are separated, thus generates a decoupled decision that may cause a performance loss. Optimization-based techniques are less used for this problem, although they can significantly improve the performance of HPC systems at the expense of higher computation time. Nowadays, HPC systems are being used for modern applications, such as big data analytics and predictive model building, that employ, in general, many short jobs. However, this information is unknown at dispatching time, and job dispatchers need to process large numbers of them quickly while ensuring high Quality-of-Service (QoS) levels. Constraint Programming (CP) has been shown to be an effective approach to tackle job dispatching problems. However, state-of-the-art CP-based job dispatchers are unable to satisfy the challenges of on-line dispatching, such as generate dispatching decisions in a brief period and integrate current and past information of the housing system. Given the previous reasons, we propose CP-based dispatchers that are more suitable for HPC systems running modern applications, generating on-line dispatching decisions in a proper time and are able to make effective use of job duration predictions to improve QoS levels, especially for workloads dominated by short jobs.

A data-driven analytics approach in the study of pneumonia’s fatalities

As huge amounts of data become available in organizations and society, specific data analytics skills and techniques are needed to explore this data and extract from it useful patterns, tendencies, models or other useful knowledge, which could be used to support the decision-making process, to define new strategies or to understand what is happening in a specific field. Only with a deep understanding of a phenomenon it is possible to fight it. In this paper, a data-driven analytics approach is used for the analysis of the increasing incidence of fatalities by pneumonia in the Portuguese population, characterizing the disease and its incidence in terms of fatalities, knowledge that can be used to define appropriate strategies that can aim to reduce this phenomenon, which has increased more than 65% in a decade.

Visual analytics for behavioral and niche market segmentation

Companies require information in order to gain an improved understanding of their customers. Data concerning customers, their interests and behavior are collected through different loyalty programs. The amount of data stored in company data bases has increased exponentially over the years and become difficult to handle. This research area is the subject of much current interest, not only in academia but also in practice, as is shown by several magazines and blogs that are covering topics on how to get to know your customers, Big Data, information visualization, and data warehousing. In this Ph.D. thesis, the Self-Organizing Map and two extensions of it – the Weighted Self-Organizing Map (WSOM) and the Self-Organizing Time Map (SOTM) – are used as data mining methods for extracting information from large amounts of customer data. The thesis focuses on how data mining methods can be used to model and analyze customer data in order to gain an overview of the customer base, as well as, for analyzing niche-markets. The thesis uses real world customer data to create models for customer profiling. Evaluation of the built models is performed by CRM experts from the retailing industry. The experts considered the information gained with help of the models to be valuable and useful for decision making and for making strategic planning for the future.

Action Research and Learning Analytics in Higher Education

Teaching is a dynamic activity. It can be very effective, if its impact is constantly monitored and adjusted to the demands of changing social contexts and needs of learners. This implies that teachers need to be aware about teaching and learning processes. Moreover, they should constantly question their didactical methods and the learning resources, which they provide to their students. They should reflect if their actions are suitable, and they should regulate their teaching, e.g., by updating learning materials based on new knowledge about learners, or by motivating learners to engage in further learning activities. In the last years, a rising interest in ‘learning analytics’ is observable. This interest is motivated by the availability of massive amounts of educational data. Also, the continuously increasing processing power, and a strong motivation for discovering new information from these pools of educational data, is pushing further developments within the learning analytics research field. Learning analytics could be a method for reflective teaching practice that enables and guides teachers to investigate and evaluate their work in future learning scenarios. However, this potentially positive impact has not yet been sufficiently verified by learning analytics research. Another method that pursues these goals is ‘action research’. Learning analytics promises to initiate action research processes because it facilitates awareness, reflection and regulation of teaching activities analogous to action research. Therefore, this thesis joins both concepts, in order to improve the design of learning analytics tools. Central research question of this thesis are: What are the dimensions of learning analytics in relation to action research, which need to be considered when designing a learning analytics tool? How does a learning analytics dashboard impact the teachers of technology-enhanced university lectures regarding ‘awareness’, ‘reflection’ and ‘action’? Does it initiate action research? Which are central requirements for a learning analytics tool, which pursues such effects? This project followed design-based research principles, in order to answer these research questions. The main contributions are: a theoretical reference model that connects action research and learning analytics, the conceptualization and implementation of a learning analytics tool, a requirements catalogue for useful and usable learning analytics design based on evaluations, a tested procedure for impact analysis, and guidelines for the introduction of learning analytics into higher education.

A telecom analytics framework for dynamic quality of service management

Since the beginning of Internet, Internet Service Providers (ISP) have seen the need of giving to users? traffic different treatments defined by agree- ments between ISP and customers. This procedure, known as Quality of Service Management, has not much changed in the last years (DiffServ and Deep Pack-et Inspection have been the most chosen mechanisms). However, the incremen-tal growth of Internet users and services jointly with the application of recent Ma- chine Learning techniques, open up the possibility of going one step for-ward in the smart management of network traffic. In this paper, we first make a survey of current tools and techniques for QoS Management. Then we intro-duce clustering and classifying Machine Learning techniques for traffic charac-terization and the concept of Quality of Experience. Finally, with all these com-ponents, we present a brand new framework that will manage in a smart way Quality of Service in a telecom Big Data based scenario, both for mobile and fixed communications.

Content-aware compression for big textual data analysis

A substantial amount of information on the Internet is present in the form of text. The value of this semi-structured and unstructured data has been widely acknowledged, with consequent scientific and commercial exploitation. The ever-increasing data production, however, pushes data analytic platforms to their limit. This thesis proposes techniques for more efficient textual big data analysis suitable for the Hadoop analytic platform. This research explores the direct processing of compressed textual data. The focus is on developing novel compression methods with a number of desirable properties to support text-based big data analysis in distributed environments. The novel contributions of this work include the following. Firstly, a Content-aware Partial Compression (CaPC) scheme is developed. CaPC makes a distinction between informational and functional content in which only the informational content is compressed. Thus, the compressed data is made transparent to existing software libraries which often rely on functional content to work. Secondly, a context-free bit-oriented compression scheme (Approximated Huffman Compression) based on the Huffman algorithm is developed. This uses a hybrid data structure that allows pattern searching in compressed data in linear time. Thirdly, several modern compression schemes have been extended so that the compressed data can be safely split with respect to logical data records in distributed file systems. Furthermore, an innovative two layer compression architecture is used, in which each compression layer is appropriate for the corresponding stage of data processing. Peripheral libraries are developed that seamlessly link the proposed compression schemes to existing analytic platforms and computational frameworks, and also make the use of the compressed data transparent to developers. The compression schemes have been evaluated for a number of standard MapReduce analysis tasks using a collection of real-world datasets. In comparison with existing solutions, they have shown substantial improvement in performance and significant reduction in system resource requirements.

Data-analytiikan hyödyntäminen yrityksen talouden ohjauksessa

Kandidaatintyö on toteutettu kirjallisuuskatsauksena, jonka tavoitteena on selvittää data-analytiikan käyttökohteita ja datan hyödyntämisen vaikutusta liiketoimintaan. Työ käsittelee data-analytiikan käyttöä ja datan tehokkaan hyödyntämisen haasteita. Työ on rajattu tarkastelemaan yrityksen talouden ohjausta, jossa analytiikkaa käytetään johdon ja rahoituksen laskentatoimessa. Datan määrän eksponentiaalinen kasvunopeus luo data-analytiikan käytölle uusia haasteita ja mahdollisuuksia. Datalla itsessään ei kuitenkaan ole suurta arvoa yritykselle, vaan arvo syntyy prosessoinnin kautta. Vaikka data-analytiikkaa tutkitaan ja käytetään jo runsaasti, se tarjoaa paljon nykyisiä sovelluksia suurempia mahdollisuuksia. Yksi työn keskeisimmistä tuloksista on, että data-analytiikalla voidaan tehostaa johdon laskentatoimea ja helpottaa rahoituksen laskentatoimen tehtäviä. Tarjolla olevan datan määrä kasvaa kuitenkin niin nopeasti, että käytettävissä oleva teknologia ja osaamisen taso eivät pysy kehityksessä mukana. Varsinkin big datan laajempi käyttöönotto ja sen tehokas hyödyntäminen vaikuttavat jatkossa talouden ohjauksen käytäntöihin ja sovelluksiin yhä enemmän.

Real-time analytics on large dynamic graphs

In today's fast-paced and interconnected digital world, the data generated by an increasing number of applications is being modeled as dynamic graphs. The graph structure encodes relationships among data items, while the structural changes to the graphs as well as the continuous stream of information produced by the entities in these graphs make them dynamic in nature. Examples include social networks where users post status updates, images, videos, etc.; phone call networks where nodes may send text messages or place phone calls; road traffic networks where the traffic behavior of the road segments changes constantly, and so on. There is a tremendous value in storing, managing, and analyzing such dynamic graphs and deriving meaningful insights in real-time. However, a majority of the work in graph analytics assumes a static setting, and there is a lack of systematic study of the various dynamic scenarios, the complexity they impose on the analysis tasks, and the challenges in building efficient systems that can support such tasks at a large scale. In this dissertation, I design a unified streaming graph data management framework, and develop prototype systems to support increasingly complex tasks on dynamic graphs. In the first part, I focus on the management and querying of distributed graph data. I develop a hybrid replication policy that monitors the read-write frequencies of the nodes to decide dynamically what data to replicate, and whether to do eager or lazy replication in order to minimize network communication and support low-latency querying. In the second part, I study parallel execution of continuous neighborhood-driven aggregates, where each node aggregates the information generated in its neighborhoods. I build my system around the notion of an aggregation overlay graph, a pre-compiled data structure that enables sharing of partial aggregates across different queries, and also allows partial pre-computation of the aggregates to minimize the query latencies and increase throughput. Finally, I extend the framework to support continuous detection and analysis of activity-based subgraphs, where subgraphs could be specified using both graph structure as well as activity conditions on the nodes. The query specification tasks in my system are expressed using a set of active structural primitives, which allows the query evaluator to use a set of novel optimization techniques, thereby achieving high throughput. Overall, in this dissertation, I define and investigate a set of novel tasks on dynamic graphs, design scalable optimization techniques, build prototype systems, and show the effectiveness of the proposed techniques through extensive evaluation using large-scale real and synthetic datasets.

Il passaggio dalla Business Intelligence all'Advanced Analytics: individuazione e analisi di un modello in merito all'influenza del lobbying nel sistema economico

I Big Data stanno guidando una rivoluzione globale. In tutti i settori, pubblici o privati, e le industrie quali Vendita al dettaglio, Sanità, Media e Trasporti, i Big Data stanno influenzando la vita di miliardi di persone. L’impatto dei Big Data è sostanziale, ma così discreto da passare inosservato alla maggior parte delle persone. Le applicazioni di Business Intelligence e Advanced Analytics vogliono studiare e trarre informazioni dai Big Data. Si studia il passaggio dalla prima alla seconda, mettendo in evidenza aspetti simili e differenze.

Analytics and Methods for Logistic System Design and Operations Management

Analytics is the technology working with the manipulation of data to produce information able to change the world we live every day. Analytics have been largely used within the last decade to cluster people’s behaviour to predict their preferences of items to buy, music to listen, movies to watch and even electoral preference. The most advanced companies succeded in controlling people’s behaviour using analytics. Despite the evidence of the super-power of analytics, they are rarely applied to the big data collected within supply chain systems (i.e. distribution network, storage systems and production plants). This PhD thesis explores the fourth research paradigm (i.e. the generation of knowledge from data) applied to supply chain system design and operations management. An ontology defining the entities and the metrics of supply chain systems is used to design data structures for data collection in supply chain systems. The consistency of this data is provided by mathematical demonstrations inspired by the factory physics theory. The availability, quantity and quality of the data within these data structures define different decision patterns. Ten decision patterns are identified, and validated on-field, to address ten different class of design and control problems in the field of supply chain systems research.

On the cloud deployment of a session abstraction for service/data aggregation

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Evaluation and optimization of a session-based middleware for data management

Dissertação para obtenção do Grau de Mestre em Engenharia Informática